Communication systems using an Internet Protocol (IP) network, etc. are known. In such a communication system, users access a server to receive various types of services, such as communication services, e.g., Voice over Internet Protocol (VoIP), and website browsing services.
In such a communication system, if communications with or accesses to a server are concentrated at one particular time, network congestion may occur. Due to such network congestion, it may take a longer time to return responses from the server, which may impair stable provision of services, such as communication services and website browsing services.
In this case, network congestion may be avoided by, for example, distributing a processing load over a plurality of pieces of hardware (servers) in advance. With the distribution of the processing load, the concentration of processing on one server occurs less frequently, thereby making it possible to stably provide services.
A technique for distributing a processing load over a plurality of servers is, for example, as follows. When a user uses a Domain Name System (DNS) server by specifying a proxy server, the DNS server may allocate, from a group of servers which are set in a load distribution environment, a server which is operating with a smallest amount of load. An example of the related art is disclosed in Japanese Laid-Open Patent Publication No. 2000-112908.